A Personal Health Monitoring and Emergency Assistance Mobile Application Jonathan D. Rosario Colón Master of Engineering in Computer Engineering Advisor: Jeffrey Duffany, Ph.D. Electrical and Computer Engineering and Computer Science Department Polytechnic University of Puerto Rico Abstract  As digital technology continues to The Be Safe mobile application aims to address advance; new opportunities arise for enhancing these challenges by leveraging cutting-edge AI emergency response and health management. The capabilities and digital tools to optimize emergency Be Safe mobile application focuses on identifying response services and provide users with real-time these opportunities and leveraging today's AI assistance during critical situations. By providing capabilities to improve the efficiency and accurate information to first responders, the Be effectiveness of emergency services. By utilizing Safe app can help reduce the response times, which, Dart, FlutterFlow, Firebase, and OpenAI's according to the American Heart Association, is ChatGPT, the app enables users to store and crucial as the chances of survival for cardiac arrest monitor vital health data, record and upload victims decrease by 7-10% with every minute delay emergency videos, and receive real-time AI-driven in defibrillation. guidance. The primary goal of the project is to In Puerto Rico, emergency calls to 911 often optimize emergency response in Puerto Rico by face issues related to response times and the providing accurate information to first responders, allocation of appropriate resources. As a result, ultimately saving time and resources. With the there is a growing need to develop solutions that integration of advanced AI-driven communication can streamline the process and provide accurate features and personalized content generation, Be information to first responders, ensuring that they Safe aims to be a powerful tool for both individuals are well-equipped to handle emergencies and first responders in managing health and effectively. With the integration of advanced AI- ensuring safety during emergencies. driven communication features and personalized Key Terms  Be Emergency Response, Health content generation, Be Safe aims to be a powerful Management, AI-driven Communication, tool for both individuals and first responders in Personalized Content, Opportunities. managing health and ensuring safety during emergencies. INTRODUCTION The Be Safe mobile application focuses on addressing these challenges by leveraging today's The increasing reliance on technology has AI capabilities to enhance emergency response revealed numerous opportunities to improve services and provide real-time guidance to users emergency response and health management, during emergencies. By utilizing Dart, FlutterFlow, particularly in the context of handling emergency Firebase, and OpenAI's ChatGPT, the app enables calls in Puerto Rico. According to the Puerto Rico users to store and monitor vital health data, record Emergency Management and Disaster and upload emergency videos, and receive real-time Administration Agency, more than 1.3 million AI-driven guidance. The primary goal of the project emergency calls were made to the 911 system in is to optimize emergency response in Puerto Rico 2019. Unfortunately, approximately 40% of these by providing accurate information to first calls were considered false alarms or non- responders, ultimately saving time and resources. emergency situations, leading to increased response As the integration of technology and times and inefficient use of resources (EMDACPR, emergency response services continues to grow, the 2019) [1]. Be Safe mobile application seeks to capitalize on these advancements to provide a comprehensive can address the challenges faced by the emergency solution for both individuals and first responders, sector. The Be Safe mobile application focuses on helping to ensure the safety and well-being of those leveraging today's AI capabilities and digital tools in need during critical situations. to optimize emergency response services and provide real-time assistance to users during BACKGROUND emergencies. With Dart, FlutterFlow, Firebase, and In Puerto Rico, emergency response systems OpenAI's ChatGPT integration, the app aims to face unique challenges due to the region's offer a user-friendly and accessible platform for geographical and infrastructural factors. The improved emergency response and health island's topography, combined with its management in everyday life. susceptibility to natural disasters such as hurricanes PROBLEM and earthquakes, highlights the need for an efficient and reliable emergency response system. As mentioned earlier, emergency response Furthermore, in a study conducted by the Puerto systems in Puerto Rico face several challenges, Rico Institute of Statistics, it was revealed that the including inefficiencies in resource allocation and a average response time for emergency medical lack of accurate information. These issues often services (EMS) in the region was 15.2 minutes in lead to delays in assistance and the deployment of 2018, significantly higher than the recommended 8- unnecessary resources, ultimately resulting in minute response time for urban areas (PRIS, 2018) wasted time and money. The Be Safe mobile [2]. application aims to address these challenges by These challenges have led to the recognition of providing a comprehensive solution that empowers the potential for technology and artificial individuals to manage their health and optimize intelligence to improve emergency response emergency response services. services. By integrating AI-driven features and A study by Rivera-Santana et al. (2016) digital tools, the Be Safe mobile application seeks highlighted the need for improvements in pre- to enhance the efficiency of emergency response hospital care in Puerto Rico, including better and empower individuals to manage their health communication between EMS and receiving proactively. The app allows users to store and hospitals and enhanced training for emergency monitor vital health data, record and upload medical technicians [3]. emergency videos, and receive real-time AI-driven The Be Safe app seeks to solve these problems guidance during emergencies. by enabling users to record and upload videos of The Be Safe app aims to address the emergencies, providing first responders with inefficiencies in resource allocation and valuable information to accurately assess the information accuracy in emergency response situation and dispatch the correct team. services, particularly in Puerto Rico. By providing Additionally, the app empowers users to manage first responders with accurate information and their health more effectively by providing AI- enabling users to record and upload videos of driven guidance and personalized content. By emergencies, the app helps in dispatching the addressing these challenges, the Be Safe mobile appropriate emergency response teams, saving time application aims to save time, resources, and and resources. In addition, the app offers users an potentially lives during emergency situations in AI-driven experience to assist in managing their Puerto Rico. health, promoting overall well-being and safety. As technology and AI continue to advance, there is an increasing need to develop solutions that METHODOLOGY ANALYSIS To develop the Be Safe mobile application and In order to gain a better understanding of the optimize its features, we will follow a multi-step current emergency response system in Puerto Rico process: and to identify opportunities for improvement, we 1. Assess the current emergency response system conducted an interview with an expert in the field. in Puerto Rico: This involves analyzing Our interviewee has 12 years of experience as a response times, resource allocation, first responder, working both in ambulances and communication between emergency service local emergency management offices. Their providers, and other key factors that affect the insights provided valuable context for our analysis efficiency of emergency response. and helped us to identify key areas where the Be 2. Identify opportunities for improvement: Based Safe app could potentially enhance the existing on the assessment, we will pinpoint specific system. areas where the Be Safe app can offer valuable During our interview, our expert emphasized support, such as facilitating communication the importance of reducing the time it takes for 911 between emergency responders and improving to transmit calls to local offices, as well as the time the allocation of resources. it takes for these offices to assign appropriate 3. Develop the AI-driven platform: The Be Safe providers to each case. They also highlighted the app will use AI technologies to enable users to potential benefits of providing emergency agents record and upload videos of emergencies and with real-time video footage of emergency receive personalized health guidance. The app situations, as this could help them to more will also analyze user data to improve the accurately assess the severity of each case and overall emergency response system. allocate resources more efficiently. 4. Conduct user testing: To ensure the app meets Based on these insights, we decided to analyze the needs of users and effectively addresses the real emergency response data to evaluate the challenges identified in the assessment, we will current system and determine how the Be Safe app engage in extensive user testing and gather could potentially improve emergency response feedback on the app's features and times and overall efficiency. The following Data functionality. Analysis and Results section presents our findings 5. Refine the app based on feedback: Using the and discusses the potential impact of the Be Safe insights gained from user testing, we will make app on the emergency response system in Puerto improvements to the app's design, features, and Rico. functionality to maximize its impact on the emergency response system in Puerto Rico. Table 1 Time to Transmit an Emergency Case Date Emergency Time to transmit to Local Office Time to Assign a Provider 1 5/8/23 Respiratory Distress 1:04 0:04 2 5/8/23 Car Crash 3:09 1:20 3 5/7/23 Bleeding from a leg 2:02 0:46 4 5/5/23 Urinal bleeding 3:07 8:42 5 5/3/23 Nervous breakdown 1:33 3:00 UNDERSTANDING THE CURRENT SYSTEM: increase the efficiency of emergency responses. We SKYCAD noticed that there are certain areas in the process, such as transmitting the call from 911 to the nearest As part of our research, we obtained supervised local office and assigning a provider, where delays access to SkyCAD, a computer-aided dispatch could occur. system currently used to manage emergency calls in We captured screenshots of the SkyCAD Puerto Rico. This was facilitated by our interface and its response time calculation for expertinterviewee, who provided guidance and reference, to provide a visual contrast to our Be context to understand its features and functions. Safe app interface and processes. These visuals will Our aim was to identify the strengths and be presented in the following sections. weaknesses of the existing system and draw parallels to our proposed solution, the Be Safe app. Figure 2 SkyCAD Dashboard RESULTS AND DISCUSSION We conducted an analysis of five real emergency cases to evaluate the current emergency response system in Puerto Rico and identify potential areas of improvement. We focused on the Figure 1 time it takes for a 911 call to be transmitted to the SkyCAD nearest local office and the time it takes for the SkyCAD, developed by SkyTec, is a robust office to assign a provider to the case. solution for handling emergency calls, from Findings receiving and dispatching to tracking and reporting. It includes a user interface with several components Our analysis indicates that the time to transmit that emergency service providers use to manage an emergency from 911 to the nearest local office incoming calls, allocate resources, and monitor the and the time to assign a provider vary significantly status of emergency responses in real-time. across cases. These highlights potential We carefully observed how SkyCAD inefficiencies in the current system and suggests calculates response times, as our primary focus that there is room for improvement. with the Be Safe app is to reduce these times and Figure 3 SkyCAD Part of Generated Case PDF Potential Impact of the Be Safe App and refine the app's features to maximize its impact Based on the collected data and insights from on the emergency response system. our interview with an experienced first responder, we believe that the Be Safe app has the potential to TOOLS AND TECHNOLOGIES improve emergency response times and efficiency FlutterFlow is a powerful and user-friendly app in several ways: development platform that helps to reduce • Reducing the transmission time from 911 to development time by providing pre-created widgets the nearest local office: By automatically and a visual interface for designing and building determining the user's location and uploading applications. FlutterFlow simplifies the the video and relevant information to the development process by generating code nearest local emergency office, the Be Safe app automatically for the app's user interface, using the could significantly reduce the time it takes for Flutter framework. This allows developers to focus 911 to transmit the emergency to the local on other aspects of the project while ensuring a office. responsive and visually appealing application [4]. • Facilitating faster provider assignment: Providing emergency agents with a video of the emergency scenario allows them to quickly assess the situation and assign an appropriate Figure 4 provider based on the observed details, such as Flutter Flow the severity of injuries, the number of people In addition to pre-built widgets, FlutterFlow involved, and other relevant factors. This could also allows developers to create custom widgets potentially reduce the time needed to assign a and integrate them into the application. This feature provider and ensure that the most appropriate provides more flexibility and control over the app's resources are allocated to each emergency. design and functionality, ensuring that unique • Improving overall emergency response requirements can be met. Furthermore, FlutterFlow efficiency: By streamlining communication supports the creation of custom functions, enabling between users, emergency service providers, developers to define specific behaviors and and local emergency offices, the Be Safe app interactions tailored to the app's needs. could help to optimize resource allocation and By leveraging FlutterFlow's capabilities, ensure that help is provided more quickly and including its support for custom widgets and effectively in emergency situations. functions, we were able to quickly create the user By comparing the current system, SkyCAD, interface, implement the app's features, and address with our proposed solution, the Be Safe app, we're any specific requirements unique to the Be Safe able to highlight the innovative features and app. improvements our app offers. It is important to Firebase is a powerful and versatile backend- remember that this comparison is not intended to as-a-service (BaaS) platform by Google that undermine the effectiveness of SkyCAD, but rather provides a suite of tools for app development and to illustrate how technology can continually evolve management, including Firestore, Authentication, and enhance emergency response management. Real-time Database, and Cloud Messaging [5]. In conclusion, our data analysis and results Firebase offers a scalable and robust infrastructure, suggest that the Be Safe app has the potential to which enables us to build and manage the Be Safe significantly improve the efficiency of emergency app efficiently. response in Puerto Rico. Further research and user testing will be necessary to confirm these findings as well as to provide personalized assistance during emergencies. Figure 5 Firebase Firestore is a flexible, scalable NoSQL cloud Figure 6 database that allows us to store and sync app data in Open AI real-time. In the Be Safe app, Firestore serves as the ChatGPT is a state-of-the-art natural language main database, enabling us to store user data, processing model developed by OpenAI that excels emergency videos, and other relevant information. at generating human-like text based on a given The real-time capabilities of Firestore ensure that input [6]. In the Be Safe app, we use ChatGPT to any changes made to the data are immediately offer a virtual assistant that understands and reflected in the app, enhancing the overall user responds to user queries in natural language. This experience. enables users to receive contextually relevant Firebase Authentication offers a variety of information and assistance during emergencies, sign-in methods, such as email/password, social such as first aid advice, guidance on contacting media logins, and phone number authentication. In emergency services, and other crucial support. the Be Safe app, we use Firebase Authentication to Text-to-Speech (TTS): TTS is a technology securely manage user sign-ins and access control, that converts written text into spoken words, ensuring that only authenticated users can access allowing users to receive information audibly rather their data and interact with the app. than reading it on the screen. We incorporate TTS Notifications: Firebase Cloud Messaging into the Be Safe app by using the Flutter Text-to- (FCM) is a service that allows us to send targeted Speech (flutter_tts) package, which supports notifications and messages to users of the Be Safe multiple languages, voices, and speech rates. By app. With FCM, we can send alerts, reminders, or implementing TTS, we aim to improve the other important information to users, keeping them accessibility of the app and cater to users who may informed and engaged with the app. have difficulty reading text or prefer to receive Cloud Functions: Firebase Cloud Functions information audibly [7]. enable us to write and deploy serverless functions Speech-to-Text (STT): STT is a technology that are triggered by events in our app, such as a that translates spoken language into written text, new video being uploaded, user registration, or enabling users to interact with the app using their changes to the database. In the Be Safe app, we can voice. In the Be Safe app, we implement STT using utilize Cloud Functions to integrate with ChatGPT, the Speech-to-Text (speech_to_text) package for send notifications, or automate other tasks, such as Flutter, which supports various languages and determining the nearest emergency office based on offers real-time speech recognition. Integrating user location. STT allows users to communicate with the app By utilizing these Firebase services, we can more efficiently and hands-free, which is efficiently build, manage, and scale the Be Safe particularly valuable in emergency situations where app, ensuring a secure, responsive, and feature-rich using a keyboard might not be practical [8]. experience for users. By integrating ChatGPT, TTS, and STT In the Be Safe app, we incorporate OpenAI's technologies into the Be Safe app, we aim to create ChatGPT, Text-to-Speech (TTS), and Speech-to- a seamless and efficient communication experience Text (STT) technologies to facilitate efficient and that caters to a diverse range of user needs, accurate communication between users and the app, preferences, and abilities. These technologies work together to provide an intuitive and accessible capture necessary visual evidence, and efficiently interface, enabling users to receive personalized communicate with emergency service providers. assistance and information during emergencies. In the following sections, we will delve into the intricate details of this specific UI, its various MAIN IMPLEMENTATIONS functionalities, and its interoperability with the The Be Safe mobile application encompasses underlying technologies such as Firebase and multiple pages designed to facilitate an efficient ChatGPT. We'll present the methodologies used in and user-friendly experience. It includes the Home its development, the challenges encountered, and Screen, which serves as the application's main hub, how we've overcome them to ensure the Be Safe providing quick and easy access to various features app can potentially transform emergency response such as the Patient Record and Profile pages. These management in Puerto Rico. pages are integral parts of the application, offering It's worth noting that while we focus on the personalized user data and a range of options to Emergency Call UI for the purpose of this report, navigate through the app effectively. the other features of the app contribute to its overall functionality and value proposition. They serve as a foundation for the seamless integration and operation of the Emergency Call feature, creating a comprehensive emergency response solution. Figure 7 Be Safe Home Page Figure 8 However, the cornerstone of our project lies Emergency Call UI within the Emergency Call User Interface (UI). Given its critical role in emergency situations, we The Recording Button have devoted a significant portion of our efforts The Be Safe mobile application leverages the towards the development and refinement of this UI. power of real-time video recording and sharing to This focus is due to the immediate impact it can assist in emergencies. The following code segment have on a user's ability to report emergencies, demonstrates how the app enables users to record and upload videos instantly, as well as send information to emergency responders, helping to relevant information to emergency responders: improve response times and outcomes in critical 1. Long-press event handler: The “onLongPress” situations. event is triggered when the user long-presses the button on the app's interface. This initiates the video recording and sharing process. Figure 9 2. Get the user's location: The app retrieves the Virtual Assistant Button user's current location using the The Virtual Assistant Button “getCurrentUserLocation” function, which returns a “LatLng” object representing the The Be Safe mobile application incorporates user's latitude and longitude. speech-to-text and text-to-speech functionality, 3. Record the video: The “recordVideo” function allowing users to communicate with a virtual is called to start recording a new video using assistant using voice commands. This is achieved the device's camera. The video recording will through the use of the “speech_to_text” and stop automatically after a predetermined “flutter_tts” packages, as well as the custom duration or when the user manually stops the functions provided by the app. The following steps recording. detail the implementation process: 4. Upload the video: The “uploadData” function 1. Display text transcription: The “Text” widget is used to upload the recorded video to a is used to display the transcribed speech from remote server. The video's download URL is the user. The text is styled using the stored in the “downloadUrls” list, which is “FlutterFlowTheme” and is updated later used to share the video with emergency dynamically as the user speaks. responders. 2. Virtual Assistant button: The 5. Update the UI: The user interface is updated to `FFButtonWidget` is used to create a button reflect the new video upload status, as well as with the label "Virtual assistant." When this any errors that may have occurred during the button is pressed, the following actions are process. triggered: 6. Store emergency call information: The a. Speech-to-text conversion: The “emergencyCalsCreateData1” and “actions.speechToText()” function is “emergencyCalsCreateData2” objects are called, which records the user's voice and created to store the relevant emergency call converts it into text using the data, such as the video URL and the user's “speech_to_text” package. location. These objects are then used to create a b. Send text to GPT API: After a short delay, new “EmergencyCalsRecord” document, the transcribed text is sent to the Chat GPT which is saved to the app's database. API using the “ChatGPTCall.call()” 7. Create a video call button: The “child” function. The JSON data from the API property of the “onLongPress” event handler response is stored in the contains a container that displays a video call “_model.chatResponse” object. button on the app's interface. When this button c. Check for a successful response: If the API is pressed, the video recording and sharing call is successful, the app retrieves the process is initiated. assistant's response from the JSON data and stores it in the “FFAppState().tts” This implementation ensures that users can variable. only share real-time video footage recorded at the d. Text-to-speech conversion: The moment, providing valuable and up-to-date “actions.talkToMe()” function is called, which uses the “flutter_tts” package to 3. The app records the video and uploads it to convert the assistant's response into Firebase, along with the user's data and speech, playing it back to the user. location. 3. Update the UI: The user interface is updated to 4. A Firebase function is triggered that identifies reflect any changes to the transcribed text, the the nearest emergency office based on the assistant's response, and the button's state. user's location. By integrating speech-to-text and text-to- 5. The Firebase function sends the emergency speech functionality, the Be Safe app provides video to the identified emergency office. users with a seamless and efficient way to 6. An emergency handler agent at the office communicate with a virtual assistant during views the video and assesses the situation. emergencies, potentially saving valuable time and 7. The agent identifies the type of emergency improving outcomes. (e.g., severe leg injury) and assigns a provider to respond. USE CASE 8. The agent also fills out an observation report based on the video, including details such as Here's a use case scenario presented in the exposed bone, right leg, and severe bleeding. figure 10 for the Be Safe app, demonstrating the 9. The assigned provider receives the emergency process when a user has an emergency and presses request and the observation report, giving them the “Record Video” button: a clear understanding of the user's needs and 1. The user encounters an emergency situation allowing them to prepare accordingly. (e.g., they have broken their leg, exposing the 10. The provider reaches the user's location and bone and causing severe bleeding). provides the necessary assistance. 2. At the Emergency Call UI of the Be Safe app the user presses the "Record Video" button. Figure 10 Use Case, Patient Press the Recording Button C A personalized user experience based on individual OMPETITIVE NALYSIS medical conditions and needs. In order to fully understand the advantages of By leveraging Firebase Functions, the Be Safe the Be Safe App, it is essential to compare its app can dynamically connect with ChatGPT to features with those of its competitors. Two well- provide tailored content and recommendations for known competitors in the emergency response and users depending on their specific medical personal safety market are RapidSOS and Life360. conditions, allergies, or other health-related The table 2 presents a comparison of key features concerns. This would not only improve the app's between the Be Safe App, RapidSOS, and Life360. overall usefulness but also help users receive The Be Safe App stands out from its accurate and relevant information during an competitors due to its unique features such as real- emergency. time video recording, speech-to-text, and text-to- To implement this, the app could be further speech capabilities. These features enhance the user developed to include a comprehensive user profile experience and provide a more comprehensive system where users can input their medical history, approach to emergency situations. Although conditions, and other pertinent information. Then, RapidSOS and Life360 offer integration with 911 Firebase Functions can be used to interact with the and location sharing, they do not provide the same ChatGPT API, allowing the app to access level of multimedia interaction and accessibility as contextually relevant information and provide the Be Safe App. This competitive analysis personalized guidance in response to user input. demonstrates the innovative approach and value For example, if a user suffers from a specific proposition of the Be Safe App in the emergency medical condition such as diabetes, the app could response and personal safety market. provide customized advice on managing blood Table 2 sugar levels during an emergency, in addition to the Identified Related Apps general emergency response recommendations. Feature BeSafe RapidSOS Life360 Such enhancements would not only improve Real-time video Yes No No the user experience and usefulness of the Be Safe recording app but also demonstrate the potential for AI-driven Speech-to-text Yes No No services like ChatGPT to be utilized in more Text-to-speech Yes No No personalized and context-specific ways, further Integration with Yes Yes Yes revolutionizing the field of emergency response and 911 personal safety applications. Location Yes Yes Yes sharing CONCLUSION User-friendly Yes Yes Yes The Be Safe mobile application represents a interface significant advancement in the realm of emergency response management. Harnessing the potential of FUTURE WORK modern technology, including FlutterFlow, As technology advances and the capabilities of Firebase, and OpenAI's ChatGPT, this innovative AI-driven services like ChatGPT continue to solution is poised to redefine emergency response evolve, there are numerous opportunities for experiences for the citizens of Puerto Rico. enhancing the Be Safe app in future updates. One Be Safe stands out with its unique and significant enhancement could involve the deeper thoughtful features, such as real-time video integration of ChatGPT into the application using recording and sharing, speech-to-text and text-to- Firebase Functions, allowing for a more speech functionality, and the integration of a virtual assistant. These features not only enhance user Ramos-Valencia, “Needs Assessment of the Emergency experience during critical moments but also provide Medical System in Puerto Rico,” in Health Sci. J., 2016, vol. 35, no. 2, pp. 80-84. responders with comprehensive and real-time information, thereby improving response times and [4] FlutterFlow. (n. d.). Build Flutter apps visually [Online]. Available: https://docs.flutterflow.io. potentially saving lives. While the existing version of the Be Safe app [5] Google Firebase. (2021). Firebase [Online]. Available: https://firebase.google.com/. represents a significant milestone, we are continuously striving to enhance its capabilities and [6] OpenAI 2021). ChatGPT [Online]. Available: https://chat.openai.com/. features in order to provide an even better user experience. Our future roadmap for Be Safe is [7] flutter_tts. (2021). Flutter Text-to-Speech Package [Online]. Available: https://pub.dev/packages/flutter_tts marked with continuous improvements and advancements. This includes rigorous user testing [8] speech_to_text. (2021). Speech-to-Text Package for Flutter to ensure optimal functionality, forming strategic [Online]. Available: https://pub.dev/packages/speech_ to_text. alliances with emergency services to facilitate seamless cooperation, and expanding our reach to new regions for broader impact. Moreover, we aim to stay at the forefront of technological advancements, integrating emerging technologies for enhanced user experience and capabilities. A significant enhancement on our radar is the incorporation of AI-driven personal health data analysis. This would allow us to provide users with bespoke advice during emergencies, tailored to their unique health conditions and needs. The future of Be Safe is one of constant evolution and innovation, driven by our unwavering commitment to improving emergency response outcomes. In essence, the Be Safe app is more than just an application; it's a life-saving tool powered by cutting-edge technology. It's an embodiment of the power of technological innovation to address real- world challenges and enhance the quality of life. REFERENCES [1] EMDACPR. (2019). Informe Estadístico del Sistema de Emergencias 9-1-1 en Puerto Rico [Online]. Available: https://www.emdacpr.pr.gov/documentos-y-estadisticas/ informe-estadistico-del-sistema-de-emergencias-9-1-1-en- puerto-rico. [2] PRIS. (2018). Tiempos de Respuesta de las Unidades de Emergencias Médicas en Puerto Rico [Online]. Available: https://estadisticas.pr/files/Publicaciones/EMS_Response_ Times_in_PR.pdf. [3] O. Rivera-Santana, A. Soto-Ortiz, J. Rodriguez- Maldonado, K. Rodriguez, E. Rodriguez-Nazario & G.